Genetic-load-eco-evolutionary-feedback-and-extinction-in-metapopulations

Project Abstract Habitat fragmentation can pose a significant risk to population survival by causing both demographic stochasticity and genetic drift within local populations to increase, thereby increasing genetic load. Higher load causes population numbers to decline, which reduces the efficiency of selection and further increases load, resulting in a positive feedback which may drive entire populations to extinction. Here, we investigate this eco-evolutionary feedback in a metapopulation consisting of local demes connected via migration, with individuals subject to deleterious mutation at a large number of loci. We first analyse the determinants of load under soft selection, where population sizes are fixed, and then build upon this to understand hard selection, where population sizes and load co-evolve. We show that under soft selection, very little gene flow (less than one migrant per generation) is enough to prevent fixation of deleterious alleles. By contrast, much higher levels of migration are required to mitigate load and prevent extinction when selection is hard, with critical migration thresholds for metapopulation persistence increasing sharply as the genome-wide deleterious mutation rate becomes comparable to the baseline population growth rate. Moreover, critical migration thresholds are highest if deleterious mutations have intermediate selection coefficients, but lower if alleles are predominantly recessive rather than additive (due to more efficient purging of recessive load within local populations). Our analysis is based on a combination of analytical approximations and simulations, allowing for a more comprehensive understanding of the factors influencing load and extinction in fragmented populations. This repository is the official implementation of the project described above. Layout The repository is split into two main directories named Fortran and Mathemamtica. The Fortran directory houses codes run with Fortran and contains 7 subdirectories. Six of the subdirectories are named based on general parameter values used (for example, the directory named Ks10h002 contains results run with parameter values; per locus strength of selection scaled by the carrying capacity, $Ks = 10$ and dominannce coefficient, $h = 0.02$). The last subdirectory called $\textit{differentK}$ contains results run with different values of carrying capacity, $K$. Within each subdirectory are .txt files named based on other specific parameters used and each .txt file contains several columms indicating computed statistics. Aside from the subdirectories, the Fortran directory also contains other fortran simulations (.f files) and their corresponding output (.txt) files. Again, in these cases, the name of the files indicate the parameters used for the run (e.g., sim_noLDL6000h002Km10.f indicates a simulation run assuming no LD (i.e., no linkage disequilibrium) with parameter values; number of loci, $L = 6000$, dominance coefficient, $h = 0.02$ and the strength of migration scaled by the carrying capacity, $Km = 10$). The Mathematica directory houses a mathematica notebook named Manuscript.nb which consists of the Mathematica codes for the analytical work done in the project. Software versions * Mathematica version 12.1 or later* GNU Fortran 14.1.0### 项目摘要 生境破碎化（habitat fragmentation）可通过提升局域种群（deme）内的人口随机性（demographic stochasticity）与遗传漂变（genetic drift）水平，进而增加遗传负荷（genetic load），对种群存续构成显著威胁。更高的遗传负荷会导致种群数量下降，这会降低选择效率并进一步加剧负荷累积，形成可推动整个种群走向灭绝的正反馈循环。在此，我们针对由通过迁移相互连接的局域种群组成的集合种群（metapopulation）中的这一生态-演化反馈循环展开研究，其中个体在大量基因座上携带有害突变。我们首先分析了软选择（soft selection）情境下遗传负荷的决定因素——此时种群大小固定；随后以此为基础，探究硬选择（hard selection）情境下种群大小与遗传负荷协同演化的机制。研究表明，在软选择条件下，极低的基因流（gene flow，每世代仅需不到1个迁移个体）即可阻止有害等位基因的固定。与之形成鲜明对比的是，当选择为硬选择时，需要更高水平的迁移才能缓解遗传负荷并避免种群灭绝；随着全基因组有害突变率（genome-wide deleterious mutation rate）趋近于基准种群增长率（baseline population growth rate），集合种群存续所需的临界迁移阈值（critical migration thresholds）会急剧升高。此外，当有害突变的选择系数（selection coefficients）处于中等水平时，临界迁移阈值达到最高；而若等位基因主要为隐性（recessive）而非加性（additive）时，阈值会更低——这是因为局域种群内对隐性遗传负荷的清除效率更高。本研究结合解析近似与模拟分析展开，旨在更全面地理解破碎化种群中影响遗传负荷与灭绝的各类因素。 本仓库为上述研究项目的官方实现代码库。 ### 仓库布局 本仓库分为两个主要目录，分别命名为Fortran与Mathematica。Fortran目录存放使用Fortran语言编写的代码，内含7个子目录。其中6个子目录以所用的通用参数值命名（例如，名为Ks10h002的目录对应运行参数为：以承载能力（carrying capacity, $K$）缩放的单座选择强度$Ks=10$，显性系数$h=0.02$的模拟结果）。最后一个名为$ extit{differentK}$的子目录存放使用不同承载能力$K$的模拟结果。 每个子目录内均存在以其他特定参数命名的.txt文件，每个.txt文件包含多列计算得到的统计量。除子目录外，Fortran目录还包含其他Fortran模拟代码（.f文件）及其对应的输出（.txt）文件。此类文件的文件名同样标注了本次模拟所用的参数（例如，sim_noLDL6000h002Km10.f对应一次未考虑连锁不平衡（linkage disequilibrium, LD）的模拟，其参数为：基因座数量$L=6000$，显性系数$h=0.02$，以及以承载能力缩放的迁移强度$Km=10$）。 Mathematica目录包含一个名为Manuscript.nb的Mathematica笔记本文件，内含本项目中解析分析所用的Mathematica代码。 ### 软件版本要求 * Mathematica 12.1 及以上版本 * GNU Fortran 14.1.0